Optimal desired trajectories for local area networks of drones

Mini and micro Unmanned Aerial Vehicles (UAVs) can provide a better service in commercial applications if they are used in networks. Flying Ad hoc Networks (FANETs) and Internet of Drones (IoD) are UAV networks where many UAVs belonging to different users coexist, and there are several routing protocols which have been developed to maintain communication in such networks. However, in application scenarios where only a limited number of UAV nodes are available, some communication paths cannot be established if an existing neighbour UAV node moves away and a new neighbour is not within reach. As response to this problem, LANs of Drones (LoDs) have been recently introduced to be used in these scenarios. A LoD is a star-connected relay network where the UAVs are connected serially in branches, in contrast to FANETs and IoDs. In a LoD, the movements of UAVs should be controlled in order to maintain the communication links. This thesis mainly focusses on the determination of desired trajectories of UAV nodes in a LoD when the leading UAV has to move to a given target. The trajectories were optimized to minimise the task completion time or the total energy consumed. Mathematical constraints were applied to the optimization problem in order to maintain the distances between the UAV nodes and between the closest UAV and the Ground Station (GS) so that they were less than the Wi-Fi communication distance. MATLAB simulations were used to verify the desired trajectory optimization method. Additional constraints were introduced to the optimization problem to derive the desired trajectories of UAVs in the presence of obstacles and image processing techniques were used to derive constraints of a realistic obstacle using an image of an obstacle. In addition, a secondary outcome of this research study was the determination of the minimum number of UAVs required to perform a given task and the minimum number of UAVs was determined for four categories of tasks. Finally, a purpose of this research was to determine the number of branches required in a LoD to obtain a given level of reliability. The reliability of a LoD was estimated by varying the number of branches and number of UAVs in each branch.